Hello Wolfram,
Since I have not received any responses on the list (that I am aware
of), please allow me to follow up and ask if you have any reservations
or comments regarding the patch set.
Thank you in advance
On 06/27/2017 10:58 PM, Joshua Frkuska wrote:
Purpose:
Its purpose is to introduce i2c slave with multimaster capabilities to the imx i2c controller driver.
This patch was tested on i.MX6q sabresd, sabreai and UDOO boards. It is a continuation of the work started in the following thread https://www.spinics.net/lists/linux-i2c/msg27340.html
The current version is version 5 and is a continuation of the discussion in http://www.spinics.net/lists/linux-i2c/msg27563.html
v5 change summary:
1. rebased patchset on i2c/for-next 5ff37d1a67e2fed0cae537ad682abb7f6647cca4 "Merge branch 'i2c/for-4.13' into i2c/for-next"
v4 change summary:
1. preserved Maxim's authorship, added rework note with signoff
2. re-arranged #include ordering
3. fixed check-patch warnings
4. redefined MAX_EVENTS to an integer
5. removed introduction of error codes
6. changed last_error (atomic_t) to type int
7. removed white lines
8. fixed multiple parenthesis alignment issues
9. removed explicit casting
10. replace udelay with usleep_range
11. removed multiple excess spacing issues
12. check for return of wait_event_interruptible_timeout
13. simplified conditional statements
14. removed dubious pinctrl handling
15. updated commit message
Testing: (validated locally with 3 different imx6q devices) For the purpose of this test, any 2 imx6 boards were hooked together to form a multimaster bus. In this configuration, slave and multimaster configurations can alternatively be stress tested. It is rebased and tested on the i2c/for-next branch.
1. enable CONFIG_I2C_SLAVE=y
2. enable CONFIG_I2C_SLAVE_EEPROM=y[m]
3. enable CONFIG_I2C_CHARDEV=y[m]
4. build the kernel
5. install the kernel/drivers on 2 imx devices
6. wire the i2c busses of both devices together
7. load kernel modules if needed
8. instantiate a slave eeprom on device 1 with address A on whatever bus it corresponds to (e.g. `echo slave-24c02 0x1066 > /sys/bus/i2c/devices/i2c-0/new_device`)
9. instantiate a slave eeprom on device 2 with address B on whatever bus it corresponds to (e.g. `echo slave-24c02 0x1064 > /sys/bus/i2c/devices/i2c-2/new_device`)
10. using i2cget/set on the appropriate bus, randomly read/write on device 1 from/to address B. (e.g. `i2cget -y 0 0x64` OR `i2cset -y 0 0x64 0x01 5`)
11. using i2cget/set on the appropriate bus, randomly read/write on device 2 from/to address A. (e.g. `i2cget -y 2 0x66` OR `i2cset -y 2 0x66 0x02 7`)
12. confirm operation between devices by observing the input and output of the i2cget/set functions as well as dumping the contents of the eeprom via the sysfs entry (/sys/bus/i2c/devices/i2c-<BUS #>/new_device)
Any constructive comments would be greatly appreciated.
Thank you
Joshua Frkuska (1):
ARM: imx_v6_v7_defconfig: Test imx i2c slave support
Maxim Syrchin (1):
i2c: imx: add slave support
arch/arm/configs/imx_v6_v7_defconfig | 7 +-
drivers/i2c/busses/i2c-imx.c | 730 +++++++++++++++++++++++++++++++++--
2 files changed, 695 insertions(+), 42 deletions(-)